Enabling Personalized Therapy with PBPK Modeling for Pediatric Epilepsy

Our partner was challenged with how to best recommend the doses for pediatric patients with GRIN-related NDD but not delay the start of an initial efficacy study.

There was also a need to ensure that the doses selected for the clinical study in a vulnerable pediatric patient population achieved exposures required for efficacy while remaining below the limits established in preclinical studies.

As part of a previous project in this collaboration, a PBPK model for radiprodil was developed to predict doses for infantile spasms, another form of severe epilepsy. This model was repurposed, to support the unique challenges of dosing infants with difficult to treat epilepsy syndrome. Given the inherent variability in dosing drugs in children, it was key to accurately estimate dose at an individual patient level. We devised a process to individualize dose escalation based on the measured individual patient exposures and the dedicated PBPK model. After the first starting dose that was the same for all patients, individual clinical patient exposure data were sent to the Global Certara PBPK modeling team. These data were fed into the model, which was refined as necessary, and “virtual-twin” simulations were performed in real time to estimate dose escalation options for that patient. The model was used to simulate not only PK but also the expected NR2B-NMDA receptor occupancy, allowing selection of dose levels predicted to achieve optimal receptor occupancy while remaining within limits established by preclinical studies. Leveraging the mechanistic approach of PBPK modeling within the Simcyp™ modeling platform allowed us to tackle nonlinearities and other complexities, enabling individualized dosing for this patient group. Our global team was positioned to handle processing and delivery of data to ensure a rapid turnaround of next dosing levels.

Certara developed the first PBPK model to predict accurate radiprodil dosing for infants. The model was initially developed in adults but was used to predict pharmacokinetics in pediatric subjects. More recently, we have incorporated new data and have leveraged the model to predict exposure and receptor occupancy in pediatric patients with GRIN-related NDD, helping to accelerate development in this serious disease with no effective treatments. A personalized approach where the PBPK model is used to set doses based on individually measured exposure has allowed GRIN Therapeutics to target the expected efficacious exposure range and stay within the exposure limits set for the ongoing study in GRIN-related NDD. Unique aspects of dosing and receptor occupancy were taken into consideration. The PBPK modeling approach enabled the selection of doses with the potential to achieve an optimal safety and efficacy profile for continued development in a pivotal Phase 3 study.

Key considerations in this development effort were the target population of children with GRIN-related NDD, a vulnerable population with several concomitant antiseizure medications, and the goal to stay within the expected therapeutic window. Given these factors, an individualized approach was necessary to have the best chance of achieving target levels of efficacy and accelerate development. This unique application of the PBPK model allowed us to generate individualized “virtual-twin” predictions of exposure in pediatric patients. Insights from our long-term partnership positioned us to re-use the available PBPK model to identify dosing levels that offered the best chance of being safe and efficacious.

With the data provided by GRIN’s bioanalytical partners, Certara’s PBPK team performed the modeling in real time, providing each patient with a predicted dose for the next escalation in less than 24 hours. Our teams based in the US, UK, Europe and other global locations collaborated seamlessly for fast turnaround and delivery of results.

Certara’s PBPK team built a reliable PBPK model for radiprodil that has many different applications and they continue to support the ongoing trials in pediatric patients. The PBPK model we established was used to support dosing recommendations in GRIN Therapeutic’s Ph1b study for GRIN-related disorders in children. Based on the known properties of radiprodil, use of the PBPK model demonstrated significant advantages compared to an empirical model. Our global team set up allowed lightning-fast turnaround in data delivery, supporting real-time dosing decision-making. As GRIN Therapeutics recently shared in the news announcement of their Phase 1b trial, the “results are expected to support planned discussions with regulators to potentially advance radiprodil to a Phase 3 pivotal trial.”